Transcript Sweden

Improving quality, accessibility
and security of biobanks using
biobank registries
Joakim Dillner
Coordinator, Swedish National Biobanking Programme
Dept of Laboratory Medicine; Lund University
Malmö University Hospital; Sweden
Sweden
Why biobanking in cancer research?
• 1. Reliability
– Complete within-cohort attendance rate. Minimised selection
biases
– Minimisation of reverse causality biases
• 2. Efficiency
– Samples and data already collected. Follow-up studies with
decades of follow-up completed in weeks
• 3. Past exposures
– Particularly for cancer, causal exposure may have occurred
long before diagnosis (even at the fetal stage) and may not be
detectable unless samples are still stored from that time.
Biobanking & Registries: An example
of follow-up from womb to tomb:
Childhood Leukemia etiology
• Sera from 403 pregnancies where the baby
developed childhood leukaemia and samples
from 1216 control pregnancies. Maternal IgM
to Epstein-Barr virus 2.9-fold RR (p<0.01).
American Journal of Epidemiology. 158: 207213 (2003)
• Analyses of other candidate agents ongoing
• Second study with new linkages to investigate
consistency planned.
Specific example of health
benefits of biobank-based
research
•
Papillomavirus (HPV) as a cause of cervical cancer
– Normal cervical smears taken up to 26 years before
cervical cancer were analysed for HPV. Strongly increased
relative risk to develop cervical cancer (16 times) (Wallin
et al, NEJM, 1999).
– All cervical smears had been saved, all could still be
located.
– Press Conference 23/5 1995- Coverage in major regional
newspapers (Västerbottenskuriren; Västerbottens
Folkblad). No woman reported „No Thank You“ to us: All
samples could be analysed.
HPV in other anogenital cancers
• HPV DNA regularly found in cancers of the
penis, anus, vulva, vagina and in
oropharyngeal cancers.
• To be able to study these more rare cancers
forms, there were not enough samples in
accessible Swedish biobanks.
The Nordic Biological Specimen Banks
working group on Cancer Causes & Control
(NBSBCCC)
• Collaborative network between 5 cancer registries and
16 different biobanks in the Nordic countries, funded by
the Nordic Council of Ministers, EU BIOMED 5 and EU
sixth framework program.
• Contains >4 million biobank specimens taken before
cancer.
• About 100.000 cases of cancer have occurred after
donation of specimen
• Code-keeping and statistical analyses at external
institutes (Cancer Registry of Norway or Finland)
Nordic biobank results on HPV and
cancer
• Series of joint Nordic seroepidemiological
studies found prospective evidence linking
HPV to these cancers
– Orofaryngeal cancer (Epithelial tonsillar cancer)
and cancers of the base of the tongue (Mork et al,
New Engl.J. Medicine 2001)
– Vulvar cancer (Bjorge et al, BMJ 1997)
– Vaginal cancer (Bjorge et al, BMJ 1997)
– Penis cancer (Bjorge et al, BMJ 1997)
– Anal cancer (Bjorge et al, BrJCa 2003)
Vaccination against
HPV
Virus-like particles: Immunologically and morphologically
identical to virus
No viral DNA- good safety profile.
Readily manufactured.
Known to completely protect against infection in animal systems
(cow, rabbit, dog monkey)
Known already 1994! Why don´t we have a vaccine yet???
Manufacturing and clinical testing of a vaccine costs billions (sic!).
Who will take this risk if the causal association is not really
sure? Prospective epidemiological data a requirement for
causality inference
Cancers caused by HPV
470.000 cases of cervical cancer (220.000 deaths) every
year.
Also high mortality in other HPV-associated cancers:
Penis, Anus, Female Genitals, certain Oral Cavity
cancers. WHO estimate: overall 5.6 % of all cancers.
Prospective evidence (required for causality inference) for
a link between HPV infection and these forms of
invasive cancer were found in a series of joint Nordic
biobank-based studies.
Vaccines now developed and found highly effective
(Koutsky et al, 2002; Harper et al 2004; Villa et al, 2005)
A Genetic riskfactor for
cervical cancer
• Twin studies have established that a genetic risk
factor for cervical cancer exists.
• Large variability in different studies of genes and
cervical cancer (Significant non-attendance of cases
and/or controls).
• Only one gene (HLA DR15/DQ6) consistently
associated, but metaanalysis finds risk increase of
only 1.47 times. Other genes more important?
• Swedish biobank-based study with complete
attendance (Ghaderi et al, 2002): 4.33 times increased
risk! Explains completely the genetic contribution to
cervical cancer.
Chlamydia trachomatis as cofactor
to HPV in cervical carcinogenesis
Biobank-based study, 26 years of follow-up, 100%
attendance: Strongly increased cancer risk for
women with Chlamydia positive smars (17
times) (Wallin et al, IJC 2002)
No increased risk in samples taken less than 5
years before cancer! Association impossible to
detect unless biobanks had been used.
Chlamydia later shown to affect whether an HPV
infection is cleared or persists (Silins et al, IJC
2005).
Nordic Biobanks and etiology of
cervical and other anogenital
cancers
• Etiology of these cancers now well known,
including interaction between genes and
environment.
• Preventable risk factors clearly identified
and prevention now well under way for
major forms of cancer.
• Real-life example of health benefits of
research based on biobanks.
Why not biobanking?
• Risks for violation of personal integrity
– Formal 3d party code-keeping. Secure databases
• Costly, logistically difficult, too much time for followup
– Ensure access and quality of existing biobanks with
long follow-up instead of starting from scratch
• Accessory information inadequate
– Assess information. Obtain information by registry
linkages
• Unreliable case ascertainment
– Don´t do biobanking in countries without
comprehensive health registers.
The Swedish National Biobanking
Programme
Sweden
• Joint national programme by the universities in Sweden-within
“Swedish FUGE” (Swegene/WCN)-funded by the Wallenberg
foundation (biggest research funding organisation in Sweden)..
Budget 2002-2005 of 54 million SEK (about 5 million euro).
• Aims:
– Improving the knowledge about the Swedish biobanking
system
– Improving the quality of the Swedish biobanks
– Increasing the usefulness and accessibility of Swedish biobanks
for health-related research as well as for clinical care and
treatment
– Improving the systems and standards for ensuring protection
of the integrity of individual donors, including promoting
ethical awareness.
QUALITY:
DESIGN AND IMPLEMENTATION OF A FORMAL
QUALITY ASSURANCE SYSTEM FOR
BIOBANKING: “Good Biobanking Practise”
• Separate, readily identifiable quality manual.
• Identity, ownership and organisation must be unambiguously
documented.
• Project management must be unambiguous, from the procedure
for scientific review of requests for samples, over integrity-assured
handling of samples and data to code-keeping and statistical
analysis.
• Security systems for preventing samples being destroyed and for
prevention of inappropriate access to samples or information
linked to samples must be in place.
• Number and type of samples must be registered.
• Standardised characterisation of usefulness of samples.
• Handling procedures to ensure maximal usefulness of samples.
Promoting usefulness, accessibility and
saving material: Sample Arrays
• Assembly of specimens and collecting accessory
information (surivival, clinical characteristics et c) is a
rate-limiting step.
• Actual testing of the specimens is another rate-limiting
step, that also uses up the materials quickly
• Tissue Arrays: Assemble population-based,
characterised tissue arrays with 1000 specimens/slide of
major diseases and of non-diseased tissue.
• DNA Arrays: Assembles purified DNA samples in
arrayed format (96 well plates) ready for genotyping
• Biobank Registry will collect corresponding longitudinal
information, e.g. on mortality, and keep the code.
Promoting scientific use of biobank samples
• SNBP: About 180 scientific papers 2002-2004.
Some specific exemples (JD)
• Prospective study on HPV in anal cancer, BrJCa 2003
• Primary EBV infection in childhood leukemia, Am J
Epidemiol, 2003
• HLA haplotypes and cervical carcinoma, IntJCancer, 2003
• Killer Immunoglobulin-like Receptor genotypes and Cervical
Carcinoma, Tissue Antigens 2005
• Vitamin D and Prostate Cancer- IntJCancer 2004
• Testosterone and Prostate Cancer IntJCancer 2004
• Enterolactone and Prostate Cancer IntJCancer 2004
• STDs and Prostate Cancer, CEBP, 2004
• Epstein-Barr virus in Multiple Sclerosis, Neurology 2004
Scientific achievements of SNBP
Some specific exemples , continued
• Vegetables and breast cancer. JAMA. 2005
• Food fibres and colon cancer Lancet 2003.
• CCP autoantibodies and HLA in rheumatoid
arthritis. Arthritis Res 2004
• Cyclin D1 and tamoxifen response. BrJCa 2004
• Use of archival plasma in large-scale
genotyping. CEBP 2005
• 15 publications on ethics and law of
biobanking
Regional Biobank Registries
• Health Care Biobanks: Enormous amount of samples
registered on PIN-identifiable files.
• Overview lacking. Access rules often unclear.
Sometimes insufficient accessory information.
Integrity issues.
• How to maximise overview?
• How to organise Quality Assurance systems?
• How to provide more information to maximise
usefulness?
• How to provide maximal integrity protection for the
individual donor in biobank-based studies?
Tasks of Regional Biobank
Registries (RGR)
• Locate and administrate destruction of sample from donor who so wishes
(Law)
• Administrate patient consent for specific uses and change of consent for
specific uses to all biobanks with samples from the donor in question (Law)
• Provide comprehensive overview over all samples that are being stored- to
the public, to the scientific community and to the health care system (all
specimen files from all health care biobanks will need to be collected)
• Coordinates implementation and control of biobank QA systems
• Collects and provides information about biobanks.
• Enable use of biospecimen for research under highest possible protection
for the individuals: External code-keeping in secure systems:
• Secure Data Base: Maximum integrity protection with multiple codings: No
research data or other information ever linked to any specific individual in
biobank-based research
First Regional Biobank Registry:
Regional Biobank
Registry of Southern Sweden
*Division within the Cancer Registry of Southern Sweden: Same
rules and mode of operation as the Cancer Registry
*Population-based registration of all biological samples stored
within a defined geographical area: Southern Sweden,
Population 1.8 million
*Jointly financed by the 4 county councils of Southern Sweden, the
University in the area (Lund University) and by the Swedish
National Biobanking program. Four employees.
*Providing files with sample ID, personal identification number,
date of sampling and type of sample to the registry is mandatory
for the health care system and also agreed with Lund University.
Retraction of consent
• Can be done at any time, also for all previously stored
samples (even for >50 year old samples).
• Send a “No-Form” specifying the type of No to the
Regional Biobank Registry (RBR).
• RBR will locate where samples are stored and ensure
that they are destroyed/not used for purposes for which
the consent has been retracted.
• Clearly defined, practical and simple possibility to
retract consent addresses some of the ethical problems
in longitudinal studies
30.000
information
leaflets per
month
distributed
for informed
consent for
saving
samples in
routine
health careincluding
consent for
research use
Informed consent statistics
april 2004-juni 2005
• Population of 1,8 million- About 300.000 informed
consents.
• 383 subjects have said ”No”.
• Biobank studies now routinely checked against the ”No
Thanks” subjects file.
• Somewhat similar system as the Danish ”Biobank No
Thank You Registry” kept at Sundhedsstyrelsen
Summary
• Biobank Registries will greatly improve
accessibility, quality and overview of
biobank samples: Ensuring maximal
usefulness for the health of the population,
while minimising integrity risks to the
individual
• Example that Biobank Sample Registration
can be implemented as a natural part of
Cancer Registration.
Acknowledgements
Swedish National Biobanking Programme: www.biobanks.se
*Wallenberg foundation
NBSBCCC:
*Nordic Council of Ministers/NordForsk
*EU Biomed 5
**EU FP6 Network of Excellence “Cancer Control using Population-based
Registries and Biobanks- CCPRB”: www.cancerbiobank.org
The Cancer Registries of Finland, Norway, Iceland, Southern Sweden,
Northern Sweden and Limburg, Belgium
German Cancer Research Center, Institute of Cancer Research (UK),
Imperial College (UK), Gliwice Oncology Center (P), Aviano Oncology
Center (I)
Twenty Participating European Biobanks (the Nordic Biobank Network
NBSBCCC and the European Network of Heriditary Cancers)